Re: Reduce the time required for a database recovery from archive. - Mailing list pgsql-hackers

From David Steele
Subject Re: Reduce the time required for a database recovery from archive.
Date
Msg-id dd7f8b17-0c87-3e2c-7a99-f972713dc227@pgmasters.net
Whole thread Raw
In response to Re: Reduce the time required for a database recovery from archive.  (Dmitry Shulga <d.shulga@postgrespro.ru>)
Responses Re: Reduce the time required for a database recovery from archive.
Re: Reduce the time required for a database recovery from archive.
List pgsql-hackers
Hi Dimtry,

On 1/11/21 2:51 AM, Dmitry Shulga wrote:
> Hi Stephen
> 
> Based on our last discussion I redesigned the implementation of WAL 
> archive recovery speed-up. 

Seems like there should have been a patch attached? In any case the 
current patch no longer applies so marked Waiting on Author.

Personally, I'm not too keen on this patch as implemented. Several 
third-party backup solutions support parallel archive get so it would be 
nice to support an interface that simply says to the restore_command, 
"go get 1gb of WAL and write the files here." This patch still assumes 
that the user has written their own restore command, which is 
third-party by definition, so I can't see how interfacing with 
third-party software is an issue here.

Also, having multiple workers blindly asking for WAL can cause quite a 
bit of traffic and cost because PG knows what WAL it wants but it 
doesn't know what exists. On the other hand, a backup solution can 
cheaply determine what is available to prevent hammering the archive 
with requests for files that don't exist.

Regards,
-- 
-David
david@pgmasters.net



pgsql-hackers by date:

Previous
From: Pavel Stehule
Date:
Subject: Re: pl/pgsql feature request: shorthand for argument and local variable references
Next
From: John Naylor
Date:
Subject: Re: Force lookahead in COPY FROM parsing